Annual Report on NCEA & New Zealand Scholarship Data & Statistics (2009)

Annual Report on NCEA & New Zealand Scholarship Data & Statistics (29) May 21

Contents Introduction Section 1 Student Achievement in NCEA and Scholarship 3 4 Internal Assessment Moderation of Internal Assessment Managing National Assessment Reports Further Assessment Opportunities 55 55 59 59 Roll Data Performance of the 27 11 cohort in NCEA and University Entrance Analyses by student gender Analyses by student ethnicity Analyses by school decile 4 8 9 12 15 Appendix A: Glossary Appendix B: Numerical data presented in figures 6 62 Performance of participating cohorts in NCEA and University Entrance (24-29) Analyses by student gender Analyses by student ethnicity Analyses by school decile 18 2 29 33 Participation data and results distributions for NQF standards 39 New Zealand Scholarship Scholarship Monetary Awards Scholarship Awards in 29 Scholarship Awards in 29 by Subject Premier Awards Premier Award and Top Subject Winners for 29 43 43 43 45 46 46 Section 2 NCEA Administrative Process and Data 47 External Assessment and the Examination Process 29 Examination Round: facts and figures The Role of NZQA in the Examination Process Marking Panels Preparing Examination Papers Assessment Schedules and judgement statements Examination Centres and Managers Personalised Examination Papers Profiles of Expected Performance Assessment Reports Reviews and Reconsiderations Breaches of the Rules 47 47 47 47 47 48 48 48 48 49 49 51 2

Introduction New Zealand s national system of assessment and qualifications for secondary school students produces a great deal of information that can be used to monitor the performance of the school system itself, and of the diverse range of students who are engaged in senior secondary education. This document, the first Annual Report on NCEA and New Zealand Scholarship Data and Statistics, summarises the activity and achievement of New Zealand s secondary school students since the full implementation of the National Certificates of Educational Achievement (NCEAs) in 24. One of the benefits of New Zealand s standards-based NCEA system of qualifications is that it allows for an analysis of the performance of students over time. In most assessments under the previous (normative) system, a candidate s final grade for an assessment was determined by the performance of that candidate relative to other candidates. In other words, normative scaling was used to ensure that the distributions of results were very similar from year to year. This scaling effectively masked any overall changes in the performance of successive cohorts. Under the NCEA system, students performance is measured relative to standards, not relative to one another. For example, if all candidates for a standard were to meet the criteria for gaining Excellence in that standard, then all would receive that grade. This means that any genuine change in performance over time will appear as a change in the achievement rate; it will no longer be hidden by scaling. In 29, the NCEA system of school qualifications saw its sixth year of full implementation. The achievement data from these six years provide an opportunity to analyse ways in which qualifications achievement by secondary students has changed since implementation. Section 1 of this report provides a series of graphs and tables showing comparisons of students qualifications achievement over time. Also in the first section, a more detailed analysis of the performance of the cohort of students that commenced 11 in 27 is presented. The data in this section allow for comparison of the performance of students grouped by gender, ethnicity and, as a proxy for their socio-economic level, the decile ratings of the schools they attend. Section 2 provides information on various aspects of internal and external assessment in schools. Also contained in this section are some administrative data; for example, on breaches of rules for external assessments. On behalf of the New Zealand Qualifications Authority (NZQA), we are pleased to make this report available to all New Zealanders. We hope that it, and subsequent reports, will contribute to public understanding of trends in senior secondary certification, of the practices and outcomes of New Zealand s senior secondary qualifications system, and of differences in the achievement of students across gender, ethnic and socio-economic categories. Bali Haque Deputy Chief Executive (Qualifications Division) Dr Michael Johnston Senior Statistician 3

Section 1 Student achievement in NCEA and Scholarship The focus of this report is on student attainment. Qualifications are therefore counted as attained provided that the requisite credit requirements have been met. This is true even if qualifications have not been formally awarded because the student has not fulfilled all administrative requirements; for example, payment of fees. The data presented in this report for 29 also exclude any successful reviews and reconsiderations, because this process was incomplete when the report was prepared. For both of these reasons, data presented in the report may differ slightly from those presented on the NZQA website. Roll Data Data and statistics pertaining to senior secondary school qualifications must be considered in relation to the particular cohorts of students engaged in attaining those qualifications. For instance, rising retention rates usually mean that more average and lower-achieving students have returned to school, which may lead to a false inference of static or declining achievement. In a standardsbased system, static achievement in an expanding cohort may actually signal an underlying improvement in achievement. This report begins with a review of trends in senior secondary rolls since the full implementation of the NCEA in 24 to provide a context for the qualifications and achievement data that follow. In the six-year period from 24 to 29, roll numbers in the senior secondary school (s 11-13) increased by some 11%, from around 144, to nearly 16,. In 29, this figure included some 6, foreign fee-paying students. Figure 1 shows that the increase varied across the years of secondary schooling, being greatest at 13, and least at 11 1. 11 12 13 7, 6, Number of students 5, 4, 3, 2, 1, 24 25 26 27 28 29 Figure 1. Numbers of students in s 11 to 13 on the school roll as of July 1, from 24 to 29. Foreign fee paying students are included. 1 This is expected: nearly 8% of year 11 students have yet to reach the age of compulsory schooling, so there is less capacity for increased retention than in s 12 and 13. 4

The greatest contribution to this increase came from a 24% rise in 13 numbers over the period. 11 and 12 numbers rose 4% and 9% respectively. Some of the increase at 13 is accounted for by a rise in the New Zealand age cohort working through the system, following relatively high birth numbers in the early 199s; an effect that peaked in 28. In addition, there has been an increase in retention from 11 to 13, especially noticeable in the declining economic environment of 29. While the number of domestic New Zealand students retained to 13 in 26 was some 59% of the 11 cohort in 24, the retention of the 27 11 cohort to 13 in 29 had risen to 66%. 13 numbers have risen faster than 11 numbers, so that in 29, 13 formed a larger proportion of the senior secondary school roll than it did in previous years, as shown in Table 1. 24 27 29 Total 11 13 Roll 136,969 148,569 153,844 11 43% 41% 4% 12 34% 34% 34% 13 23% 25% 26% Table1. Senior secondary (s 11-13) roll numbers, and 11, 12 and 13 rolls as percentages of the total senior school roll: 24 29. These include domestic New Zealand students only; foreign fee-paying students are excluded. Figure 2 shows trends in roll numbers from 24 to 29, partitioned by gender. Retention to 13 rose between 24 and 29 for both males and females. The number of 13 domestic New Zealand male students retained in 26 from the 11 cohort of two years earlier was 56%, compared to a retention figure of 63% for females. In 29, the retention of males rose to 62%, compared to 7% for females. Since 25, males have outnumbered females in 11, but the reverse has long been the case at 13 because of the higher female retention rate. 35 11 Male 12 Male 13 Male 11 Female 12 Female 13 Female 3 25 Number of students 2 15 1 5 24 25 26 27 28 29 Figure 2. Numbers of domestic New Zealand male and female students in s 11 to 13 on the school roll as at July 1 from 24 to 29. 5

Figures 3 5 show trends in roll numbers by year level from 24 to 29, partitioned by ethnicity. Between 24 and 29, the number of 11 New Zealand domestic students rose by some 17% for New Zealand Māori, and 23% for Pasifika, while New Zealand European numbers dropped slightly 2. At 13, the cohorts increased for all ethnicities: up 2% for European, 47% for Māori and 53% for Pasifika. 11 to 13 retention rates rose between 26 and 29: from 59% to 67% for European, 4% to 47% for Māori, and 67% to 71% for Pasifika. Māori now comprise 18% of senior secondary school students, up from 16% in 24, while the percentage of Pasifika rose from 8% to 9% of the senior secondary school student roll over the same period. NZ European NZ Māori Pasifika Asian Other ethnicities 45, 4, 35, Number of 11 stduents 3, 25, 2, 15, 1, 5, 24 25 26 27 28 29 NZ European NZ Māori Pasifika Asian Other ethnicities 35, 3, Number of 12 stduents 25, 2, 15, 1, 5, 24 25 26 27 28 29 2 The gradual NZ European decline evident after the peak in 26 is a population, rather than a retention effect. 6

3, NZ European NZ Māori Pasifika Asian Other ethnicities Number of 13 stduents 25, 2, 15, 1, 5, 24 25 26 27 28 29 Figures 3 5. Numbers of New Zealand European, New Zealand Māori, Pasifika, Asian and other-ethnicity domestic students in s 11 to 13 on the school roll as at July 1 from 24 to 29 3. 3 Note that numbers of Asian students are very similar to numbers of Māori students and Pasifika students, and these ethnicities therefore do not distinguish clearly in Figures 3-5. 7

Performance of the 27 11 cohort in NCEA and University Entrance The following graphs (Figures 6 13) compare the percentages of various demographic categories of students from the cohort of students commencing 11 at the beginning of 27, attaining NCEA Levels 1, 2 and 3. These are the students who, if they stayed at school, completed 13 at the end of 29. The demographic categories of interest are gender, ethnicity, and school decile, which serves as a proxy for students socioeconomic level. Comparisons between these groups of students are made for qualifications attained by the end of 27 ( 11), by the end of 28 ( 12), and by the end of 29 ( 13). Calculating percentages of students attaining qualifications over the original 11 cohort largely corrects for differences in retention between the demographic groups of interest. For example, in each year, a greater proportion of male students than female students have left school without NCEA Level 1 during s 11 and 12. Comparing the percentages of male and female students who have attained NCEA Level 1 by the end of 12 with just those students who were still at school, therefore underestimates the Level 1 performance gap between male students and female students. Using the original 11 numbers for all percentages avoids this problem because all students are counted in denominators for the percentages, whether or not they have left school. While a small number of students remain at school beyond 13, the percentage of the original 11 roll achieving a qualification by the end of 13 provides a close approximation to the proportion of school leavers with that qualification. The figures tracking the 27 11 cohort as they gained NCEA qualifications during 27, 28 and 29 compare the performance of male and female students (Figures 6-8), New Zealand European, Māori, Pasifika and Asian students (Figures 9-11) and students in low-decile (1-3), medium-decile (4-7) and high-decile (8-1) schools (Figures 12-14). Many students in New Zealand secondary schools pursue qualifications in addition to, or in lieu of, NCEA qualifications. Some of these qualifications are registered on the National Qualifications Framework; for example, the National Certificate in Computing; whereas others are not. Thus the data presented in Figures 6 14 to some extent underestimate overall qualifications achievement rates in secondary schools because they include only NCEA qualifications. The data presented in Figures 6-14 include students resident in New Zealand only; they exclude a small number of students in the Cook Islands and Niue who attain NCEA qualifications. 8

Analyses by student gender Figures 6-8 compare attainment of NCEA Level 1-3 across school s 11-13 for male and female students, and show the approximate proportions of each gender leaving school with each level of NCEA. For all three levels, these proportions are higher for female students than for male students. Figure 6 shows the percentages of male and female students commencing 11 in 27, who had attained NCEA Level 1 by the end of 27, 28 and 29. The majority of students who gained NCEA Level 1 did so in 11; approximately 57% of 11 male students and 67% of 11 female students. An approximate further 12% of the original male 11 cohort, and 1% of the original female cohort had gained Level 1 by the end of 12, with only a further two percent of the male cohort and one percent of the female cohort gaining this qualification by the end of 13. The slightly higher attainment of Level 1 by male students in s 12 and 13 diminished the difference in the cumulative attainment rate in favour of girls, from around 1 percentage points at the end of 11, to around eight percentage points at the end of 13. Male Female 1 Cumulative percentages of the 27 11 cohort achieving NCEA Level 1 8 6 4 2 27 ( 11) 28 ( 12) 29 ( 13) Calendar (School ) Figure 6. Percentages of male and female students in 11 in 27, achieving NCEA Level 1 by the end of 27 ( 11), 28 ( 12) and 29 ( 13). All standard errors are less than one percentage point. 9

Figure 7 shows the percentages of male and female students commencing 11 in 27, who had attained NCEA Level 2 by the end of 27, 28 and 29. Only around one percent of male or female students attained NCEA Level 2 prior to 12, with 6% of female students and 48% of male students attaining this qualification by the end of 12. This 12 percentage point difference in favour of female students closed somewhat by the end of the following year, with a further eight and six percent of the original 11 cohort of male and female students respectively attaining NCEA Level 2 during 13. Male Female 1 Cumulative percentages of the 27 11 cohort achieving NCEA Level 2 8 6 4 2 27 ( 11) 28 ( 12) 29 ( 13) Calendar (School ) Figure 7. Percentages of male and female students in 11 in 27, achieving NCEA Level 2 by the end of 27 ( 11), 28 ( 12) and 29 ( 13). All standard errors are less than one percentage point. 1

Figure 8 shows the percentages of male and female students commencing 11 in 27, who had attained NCEA Level 3 by the end of 27, 28 and 29. A negligible proportion of students of either gender attained this qualification during 11, and under one percent of male students and half a percent of female students attained it during 12. By the end of 13, 28% of the original 11 male cohort, and 41% of the female cohort had gained NCEA Level 3; a difference of 13 percentage points in favour of female students. Male Female 1 Cumulative percentages of the 27 11 cohort achieving NCEA Level 3 8 6 4 2 27 ( 11) 28 ( 12) 29 ( 13) Calendar (School ) Figure 8. Percentages of male and female students in 11 in 27, achieving NCEA Level 3 by the end of 27 ( 11), 28 ( 12) and 29 ( 13). All standard errors are less than one percentage point. 11

Analyses by student ethnicity Figures 9-11 compare attainment of NCEA Levels 1-3 across school s 11-13 for students who self-identify as New Zealand European, New Zealand Māori, Pasifika or Asian, and show the approximate proportions of each ethnicity leaving school with each level of NCEA. Students not identifying with any of these ethnicities are omitted from these data. Foreign fee-paying students are also excluded because ethnicity data are unavailable for these students. Figure 9 shows the percentages of New Zealand European, Māori, Pasifika and Asian students commencing 11 in 27, who had attained NCEA Level 1 by the end of 27, 28 and 29. By the end of 11 in 27, 71% of European students, 69% of Asian students 44% of Māori students and 42% of Pasifika students had attained this qualification. Thus, considerably greater percentages of European and Asian than Māori or Pasifika students had attained NCEA Level 1 by the end of 11. By the end of 12 the gaps, while still significant, had closed somewhat, with 58% of Māori, 64% of Pasifika, and around 8% of Asian and European students having attained Level 1. Pasifika students, two percentage points behind Māori students after 11, were six percentage points ahead after 12. The Pasifika cohort continued to make gains in attaining NCEA Level 1 during 13, with a further four percent of the original 11 cohort gaining the qualification. In comparison, only two percent of the Asian and Māori cohorts, and one percent of the European cohort gained NCEA Level 1 during 13. In summary, these data show that approximately 8% of Asian and European students, 68% of Pasifika students, and approximately 6% of Māori students leave school with at least NCEA Level 1. NZ European Maori Pasifika Asian 1 Cumulative percentage of the 27 11 cohort achieving NCEA Level 1 8 6 4 2 27 ( 11) 28 ( 12) 29 ( 13) Calendar (School ) Figure 9. Percentages of New Zealand European, Māori, Pasifika and Asian students in 11 in 27, achieving NCEA Level 1 by the end of 27 ( 11), 28 ( 12) and 29 ( 13). All standard errors are less than one percentage point. 12

Figure 1 shows the percentages of New Zealand European, Māori, Pasifika and Asian students commencing 11 in 27, who had attained NCEA Level 2 by the end of 27, 28 and 29. No more than two percent of any ethnic cohort attained NCEA Level 2 prior to 12. By the end of 12, a large performance difference in favour of New Zealand European (62%) and Asian (67%) students relative to Māori (35%) and Pasifika (39%) students is evident. During 13, this difference diminished, especially for Pasifika students with a further 16% of the original 11 Pasifika cohort gaining Level 2 during 13, compared with eight percent of the Māori cohort, seven percent of the Asian cohort, and five percent of the European cohort. The NCEA Level 1 shown in Figure 9 and those for Level 2 in Figure 1 show different comparative attainment of these qualifications by European and Asian students. Specifically, whereas equal proportions of European students and Asian students had attained NCEA Level 1 by the end of 13, for NCEA Level 2, the attainment rate by the end of 13 was some eight percentage points higher for Asian students than for European students. NZ European Maori Pasifika Asian 1 Cumulative percentage of the 27 11 cohort achieving NCEA Level 2 8 6 4 2 27 ( 11) 28 ( 12) 29 ( 13) Calendar (School ) Figure 1. Percentages of New Zealand European, Māori, Pasifika and Asian students in 11 in 27, achieving NCEA Level 2 by the end of 27 ( 11), 28 ( 12) and 29 ( 13). All standard errors are less than one percentage point. 13

Figure 11 shows the percentages of New Zealand European, Māori, Pasifika and Asian students commencing 11 in 27, who had attained NCEA Level 3 by the end of 27, 28 and 29. No more than one percent of any ethnic group attained NCEA Level 3 prior to 13. By the end of 13, very substantial differences between the percentages of the various ethnic groups that have attained NCEA Level 3 are evident. The percentage of Asian students gaining Level 3 by this stage was 53%, compared with 4% of European students, 16% of Māori students and 2% of Pasifika students. NZ European Maori Pasifika Asian 1 Cumulative percentage of the 27 11 cohort achieving NCEA Level 3 8 6 4 2 27 ( 11) 28 ( 12) 29 ( 13) Calendar (School ) Figure 11. Percentages of New Zealand European, Māori, Pasifika and Asian students in 11 in 27 achieving NCEA Level 3 by the end of 27 ( 11), 28 ( 12) and 29 ( 13). All standard errors are less than one percentage point. 14

Analyses by school decile Figures 12-14 explore the effects of students socioeconomic level, approximated using schools decile ratings, on performance in NCEA. An important consideration in respect of comparisons across school deciles is that this variable is confounded with ethnicity; greater proportions of students at low-decile schools than at high-decile schools identify as Māori or Pasifika. Thus it is probable that the lower average socio-economic level of Māori and Pasifika students is a factor influencing the lower qualifications attainment rates for students of these ethnicities. Data for schools without decile ratings are excluded. Figure 12 shows the percentages of students at low, medium- and high-decile schools, commencing 11 in 27, who had attained NCEA Level 1 by the end of 27, 28 and 29. Decile-related achievement differences are evident across all year levels, with 48% of students at low-decile schools, 63% of students at medium-decile schools and 74% of students at high-decile schools having attained NCEA Level 1 by the end of 11. The differences diminished by the end of 13, by which time the percentages were 66%, 77% and 83% respectively. Decile 1-3 Decile 4-7 Decile 8-1 1 Cumulative percentages of 27 11 students achieving NCEA Level 1 8 6 4 2 27 28 29 Calendar Figure 12. Percentages of students in Decile 1-3, Decile 4-7 and Decile 8-1 schools, in 11 in 27, achieving NCEA Level 1 by the end of 27 ( 11), 28 ( 12) and 29 ( 13). All standard errors are less than one percentage point. 15

Figure 13 shows the percentages of students at low, medium- and high-decile schools, commencing 11 in 27, who had attained NCEA Level 2 by the end of 27, 28 and 29. Just one percent of students at schools in each decile band attained NCEA Level 2 prior to 12. The decile-related differences in attainment of this qualification at the end of s 12 and 13 are greater than for NCEA Level 1. At the end of 12, the percentage of students attaining NCEA Level 2 at high-decile schools (67%) was close to 3 percentage points higher than the percentage at low-decile schools (39%). The percentage for students at medium-decile schools was 54%. The differences in the attainment rates across the decile bands was slightly reduced by the end of 13, being 73% for high-decile schools, 62% for medium-decile schools, and 49% for low-decile schools. Decile 1-3 Decile 4-7 Decile 8-1 1 Cumulative percentages of 27 11 students achieving NCEA Level 2 8 6 4 2 27 28 29 Calendar Figure 13. Percentages of students in Decile 1-3, Decile 4-7 and Decile 8-1 schools, in 11 in 27, achieving NCEA Level 2 by the end of 27 ( 11), 28 ( 12) and 29 ( 13). All standard errors are less than one percentage point. 16

Figure 14 shows the percentages of students at low, medium- and high-decile schools, commencing 11 in 27, who had attained NCEA Level 3 by the end of 27, 28 and 29. Less than one percent of students at schools in each decile band attained NCEA Level 3 prior to 13. By the end of 13, large differences in the rates of attainment of NCEA Level 3 were evident; at high-decile schools 46% of the original 11 cohort attained the qualification compared with 33% at medium-decile schools and 19% at low-decile schools. Decile 1-3 Decile 4-7 Decile 8-1 1 Cumulative percentages of 27 11 students achieving NCEA Level 3 8 6 4 2 27 28 29 Calendar Figure 14. Percentages of students in Decile 1-3, Decile 4-7 and Decile 8-1 schools, in 11 in 27, achieving NCEA Level 3 by the end of 27 ( 11), 28 ( 12) and 29 ( 13). All standard errors are less than one percentage point. 17

Performance of participating cohorts in NCEA and University Entrance (24-29) Table 2 below, and the following graphs (Figures 15 35) compare attainment rates over time, between genders, ethnicities and decile bands, of NCEA qualifications and University Entrance in the year most typical for gaining each; 11 for NCEA Level 1, 12 for NCEA Level 2, and 13 for NCEA Level 3 and University Entrance. Attainment is measured in terms of the percentage of the participating cohort for each qualification that gained that qualification during the typical year for doing so. A student is in the participating cohort for a given NCEA level in a given year if, on the basis of any credits they have already acquired and credits that they are entered for, it is possible for them to acquire that qualification by the end of that year. The NCEA Level 3 cohort is used as the University Entrance cohort as well. This analysis of University Entrance attainment is therefore somewhat flawed; not all students with sufficient entries to gain NCEA Level 3 have a configuration of credits that would allow them to attain University Entrance, and many students entered for sufficient credits to attain University Entrance are not entered for sufficient to gain NCEA Level 3. The participating cohort for each qualification is a proxy for students intending to attain each, there being no formal entry process for NCEA qualifications or University Entrance. If a student is not entered for sufficient credits to attain a given qualification, it is assumed they have no intention of doing so in that year. Many students pursue qualifications other than NCEA, including others registered on the National Qualifications Framework (NQF) as well as non-nqf qualifications. The percentages of students gaining a given qualification in the following figures can therefore be treated as an approximation to the proportion of students intending to gain that qualification, who did in fact attain it. Table 2 shows the proportions of participating cohorts gaining NCEA Levels 1-3 and University Entrance from 24 to 29. NCEA Levels 1 and 2 show a trend towards greater proportions of participating candidates gaining the qualification between 24 and 27. From 27 on, these proportions stabilised at around 71% for Level 1 and 76% for Level 2. The percentages of participating candidates gaining NCEA Level 3 and University Entrance similarly increased between 24 and 27. However, from 27 on, a fall in these percentages is evident. This fall might reflect greater retention of students into 13 in 28 and 29 than in previous years. A comparison between the Level 2 attainment rate in 28 from Table 2 with the data in Figure 6, and of the Level 3 attainment rate in 29 from Table 2 with the data in Figure 7, makes clear the limitations of using roll data to estimate NCEA attainment rates: the participating cohort for each NCEA level is considerably smaller than roll numbers in the typical year for attaining each level. An unsurprising corollary is that the proportion of the participating cohort gaining each level is commensurately higher than the proportion of the typical-year roll gaining each level.the attainment of qualifications other than NCEA, while beyond the scope of this report, is an important factor in explaining this discrepancy. 24 25 26 27 28 29 NCEA Level 1 ( 11) 65.6 66. 69.3 72.5 7.6 71.6 NCEA Level 2 ( 12) 72.7 72.8 74.9 76.9 75.5 75.7 NCEA Level 3 ( 13) 68. 68.4 7.5 72. 7.2 69.4 University Entrance ( 13) 64.3 64.4 66.5 67.7 65.6 64.3 Table 2. Percentages of participating cohorts gaining NCEA Level 1 in 11, NCEA Level 2 in 12, and NCEA Level 3 and University Entrance in 13. Note that the participating cohort for University Entrance is defined to be the same as that for NCEA Level 3. All standard errors are less than one percentage point. 18

Certificate endorsement for NCEA Levels 1, 2 and 3 was introduced in 27 to increase the incentive for more able students to achieve to their maximum potential. The percentage of students gaining qualifications in 29 at their typical level who received endorsements is shown in Table 3. The percentages of NCEA qualifications at each level awarded with endorsements of Merit or Excellence have been constant within one percentage point over the three years since their introduction. However, there are some variations between genders, between ethnicities and between students attending schools with different socioeconomic decile ratings 4. NCEA Level 1 11 students NCEA Level 2 12 students NCEA Level 3 13 students No endorsement 65% 75% 74% Merit endorsement 27% 19% 21% Excellence endorsement 8% 6% 5% Table 3. Percentages of Level 1, 2 and 3 NCEA qualifications gained in the typical year for each with endorsements of Merit and Excellence in 29. 4 Male/female endorsements for 27 29 are shown in Figures 22 24 for each NCEA level. Ethnicity and decile endorsement data are available in statistical tables published on the NZ Qualifications Authority website at: http://www.nzqa.govt.nz/qualifications/ssq/statistics/statsreports.do 19

Analyses by Student Gender Figures 15 18 compare the performance of male and female participating cohorts in attaining NCEA Levels 1-3 and University Entrance in the typical year for doing so, in each year since the full implementation of the NCEA qualification in 24. Figure 15 compares the percentages of male and female participating candidates gaining NCEA Level 1 between 24 and 29. A difference of seven to ten percentage points in favour of female candidates is evident in each year. The increasing trend evident in the overall data shown in Table 2 is also evident here for both genders between 24 and 27. After 27 the percentages stabilised at around 67% for males and 75% for females, although the data for male candidates have fluctuated somewhat over the last three years. Male Female 1 9 Percentage of participating 11 cohort gaining NCEA Level 1 8 7 6 5 4 3 2 1 24 25 26 27 28 29 Figure 15. Percentages of participating 11 male and female students gaining NCEA Level 1 from 24 to 29. All standard errors are less than one percentage point. 2

Figure 16 compares the percentages of male and female participating candidates gaining NCEA Level 2 between 24 and 29. There is a very consistent difference in favour of female candidates of between nine and ten percentage points in each of these years. Again, the trend for both genders closely resembles that of the overall data shown in Table 2, with a gradual increase in the proportions of candidates gaining the qualification between 24 and 27, and success rates stabilising at around 8% for female candidates and 7% for male candidates from 27 on, although the rates for male candidates in 28 and 29 were around two percentage points lower than they were in 27. Male Female 1 9 Percentage of participating 12 cohort gaining NCEA Level 2 8 7 6 5 4 3 2 1 24 25 26 27 28 29 Figure 16. Percentages of participating 12 male and female students gaining NCEA Level 2 from 24 to 29. All standard errors are less than one percentage point. 21

Figure 17 compares the percentages of male and female participating candidates gaining NCEA Level 3 between 24 and 29. Differences in favour of female candidates range between eight and ten percentage points between 24 and 27, with some tendency towards increasing after 27, predominantly because of a fall in the male attainment rate. During this period, like NCEA Levels 1 and 2, an increasing attainment rate is evident for both genders between 24 and 27. Beyond 27 however, a slight decline in the rate of success for female candidates is evident, with a more substantial decline for male candidates, resulting in a widening of the gender difference to 12 percentage points in 29. The decline is probably at least in part attributable to increased retention into 13. This increase was more pronounced for male students than for female students (see Figure 2), which might explain why the decline for males is greater than that for females. Male Female 1 9 Percentage of participating 13 cohort gaining NCEA Level 3 8 7 6 5 4 3 2 1 24 25 26 27 28 29 Figure 17. Percentages of participating 13 male and female students gaining NCEA Level 3 from 24 to 29. All standard errors are less than one percentage point. 22

Figure 18 compares the percentages of 13 male and female participants in NCEA Level 3 who gained University Entrance, in each year from 24 to 29. The pattern of attainment rates over the time period represented by the graph is very similar to that for NCEA Level 3, with differences in favour of female candidates of between six and ten percentage points. Like the data for NCEA Level 3, the success rate for University Entrance increased between 24 and 27, after which it declined somewhat for both genders and for male candidates especially. Again, the decline is likely to be influenced by increased retention into 13 (see Figure 2). An issue of interest in regard to the relative success in qualifications achievement by male and female students, is whether there is any advantage to either gender, or both, from attending either co-educational or single-sex schools. A very important consideration here is the decile of schools; single-sex schools are more highly represented amongst high-decile schools than are co-educational schools, and high-decile schools also tend to have stronger qualifications outcomes than low-decile schools. Thus, any comparison of single-sex and co-educational schools that did not take decile into account would show an advantage in qualifications achievement to single-sex schools on the basis of their higher socio-economic status alone. Figures 19-21 compare the performance during 29, of male and female participating candidates, in acquiring NCEA Levels 1, 2 and 3, and University Entrance, respectively, during the typical secondary-school year for each. The data are also partitioned by school gender (co-educational or single-sex) and by decile band. Data for students at schools without decile ratings are omitted from these comparisons. Male Female 1 9 Percentage of 13 NCEA Level 3 participants University Entrance 8 7 6 5 4 3 2 1 24 25 26 27 28 29 Figure 18. Percentages of 13 male and female NCEA Level 3 participants gaining University Entrance from 24 to 29. All standard errors are less than one percentage point. 23

Figure 19 shows the percentages of male and female participating 11 candidates gaining NCEA Level 1 during 29, at both single-sex and co-educational schools, for low, medium and high deciles respectively. In all decile bands, there is a difference in favour of female candidates, although in the low-decile band, boys at single-sex schools had a higher rate of success than girls at co-educational schools. There is a difference in favour of single-sex schools for both genders in all decile bands. The effect is most pronounced in the low-decile bands, with differences of 15 and 17 percentage points for female and male candidates respectively. The differences between the rates of success in single-sex and co-educational schools are reasonably comparable between the genders for all decile bands. Single Sex Female Co-ed Female Single Sex Male Co-ed Male 1 9 Percentage of participating 11 cohort gaining NCEA Level 1 8 7 6 5 4 3 2 1 Decile 1-3 Decile 4-7 Decile 8-1 Decile band Figure 19. Percentages of participating 11 male and female students in single-sex and co-educational schools gaining NCEA Level 1 in 29 partitioned by decile band. Error bars denote the standard errors of the proportions. 24

Figure 2 shows the percentages of male and female participating 12 candidates gaining NCEA Level 2 during 29, at low, medium- and high-decile single-sex and co-educational schools. The pattern of data is very similar to that pertaining to NCEA Level 1 presented in Figure 19 above. Again, a difference in favour of female candidates is evident in all decile bands, although in the low-decile band only, boys at single-sex schools had a higher rate of success than girls at co-educational schools. Again also, a difference in favour of single-sex schools is evident for both genders in all decile bands, with the differences particularly pronounced in the low-decile band. Single Sex Female Co-ed Female Single Sex Male Co-ed Male 1 9 Percentage of participating 12 cohort gaining NCEA Level 2 8 7 6 5 4 3 2 1 Decile 1-3 Decile 4-7 Decile 8-1 Decile band Figure 2. Percentages of participating 12 male and female students in single-sex and co-educational schools gaining NCEA Level 2 in 29 partitioned by decile band. Error bars denote the standard errors of the proportions. 25

Figure 21 shows the percentages of male and female participating candidates gaining NCEA Level 3 in 13 during 29, at low-, medium- and high-decile single-sex and co-educational schools. Again, the pattern of data is very similar to the data pertaining to NCEA Levels 1 and 2 presented in Figures 19 and 2 above: There is a difference in favour of female candidates in all decile bands, with the exception that, in the lowdecile band only, male candidates at single-sex schools had a higher rate of success than female candidates at co-educational schools. There is a difference in favour of single-sex schools for all decile bands, especially the low-decile band, and for both genders. Single Sex Female Co-ed Female Single Sex Male Co-ed Male 1 9 Percentage of participating 13 cohort gaining NCEA Level 3 8 7 6 5 4 3 2 1 Decile 1-3 Decile 4-7 Decile 8-1 Decile band Figure 21. Percentages of participating 13 male and female students in single-sex and co-educational schools gaining NCEA Level 3 in 29, partitioned by decile band. Error bars denote the standard errors of the proportions. 26

Figure 22 shows the percentages of male and female candidates participating in NCEA Level 3, and gaining University Entrance in 13 during 29, at low-, medium- and high-decile single-sex and co-educational schools. The pattern of data is similar to that for the NCEA data presented in Figures 19-21. There is a difference in favour of female candidates in all decile bands. The difference in the low-decile band in favour of male candidates at single-sex schools in comparison with female candidates at co-educational schools that was evident for all three levels of NCEA, is not statistically significant here. Again, however, there is a difference in favour of single-sex schools for all decile bands, especially the low-decile band, and for both genders. Single Sex F Co-ed F Single Sex M Co-ed M 1 Percentage of 13 participating cohort for NCEA Level 3 gaining University Entrance 9 8 7 6 5 4 3 2 1 Decile 1-3 Decile 4-7 Decile 8-1 Decile band Figure 22. Percentages of 13 male and female students participating in NCEA Level 3 in single-sex and co-educational schools gaining University Entrance in 29 partitioned by decile band. Error bars denote the standard errors of the proportions. 27

Figures 23-25 compare the percentages of male and female candidates gaining each level of NCEA who attained those qualifications with endorsements of Merit and Excellence. Significant differences in favour of female candidates are evident for both Merit and Excellence endorsements at all three levels, with the exception of Level 3 Excellence (for which the difference is not significant). A slight upward year-on-year trend in proportions gaining endorsements appears to be occurring in some cases; in particular, Excellence endorsements for both genders at all levels, and Merit endorsements for male candidates at Level 2 and 3; but continuation for a further year is needed before the significance, or otherwise, of the trend can be assessed. Percentage of 11 candidates gaining NCEA Level 1 with Merit and Excellence endorsements 35 3 25 2 15 1 5 Male Female 27 28 29 27 28 29 Merit Endorsement Excellence Endorsement Percentage of 12 candidates gaining NCEA Level 2 with Merit and Excellence endorsements 25 2 15 1 5 Male Female 27 28 29 27 28 29 Merit Endorsement Excellence Endorsement Percentage of 13 candidates gaining NCEA Level 3 with Merit and Excellence endorsements 25 2 15 1 5 Male Female 27 28 29 27 28 29 Merit Endorsement Excellence Endorsement Figures 23-25. Percentages of NCEA qualifications gained with Merit or Excellence endorsements by male and female candidates from 27 to 29 in the typical school year for each level. All standard errors are less than one percentage point. 28

Analyses by ethnicity Figures 26-29 compare the performance of Asian, New Zealand European, New Zealand Māori and Pasifika participating candidates; the four most numerous ethnic groups; in gaining NCEA Levels 1 3 and University Entrance, respectively. Students not identifying with any of these ethnicities are omitted from these data. The data in these figures should be interpreted in conjunction with those shown in Figures 3 35, which show similar comparisons across decile bands. This is because ethnic identity is correlated with socio-economic status, such that New Zealand Māori and Pasifika students are over-represented in low-decile schools. Therefore, at least some of what appears to be an ethnic effect is likely actually to be an effect of socio-economic level. Even so, a comparison of the two sets of data elucidates some effects that appear to be specifically ethnic. Specifically, the performance of both Māori and Pasifika students, especially the latter, appears to have improved between 25 and 27 considerably more than overall performance at lowdecile schools. Figure 26 compares the percentages of participating candidates in 11, gaining NCEA Level 1, across ethnic groups. New Zealand European and Asian candidates have substantially higher rates of success in gaining NCEA Level 1 during 11 than New Zealand Māori or Pasifika candidates; differences of 19-37 percentage points in favour of the former two ethnicities is evident across the time period covered by the data. There is a small difference of two to five percentage points in favour of European candidates relative to Asian candidates, and a larger difference of five to 12 percentage points in favour of Māori candidates relative to Pasifika candidates. A caveat on the latter difference is that these data are for 11 candidates only, and Pasifika candidates show a high rate of attaining NCEA Level 1 in s 12 and 13, such that, by the end of 13 a higher proportion of the original 11 Pasifika cohort had acquired NCEA Level 1 in comparison with the original 11 Māori cohort (see Figure 9). There is some evidence that the large difference between New Zealand European and Asian candidates on one hand, and New Zealand Māori and Pasifika candidates on the other, has diminished over time. Between 24 and 27, rates of success for 11 candidates in gaining NCEA Level 1 improved for all ethnicities, but especially for Māori and Pasifika. From 27 onward, the rates for all ethnicities largely stabilised. The diminution of the ethnically-based differences in attainment of NCEA Level 1 is therefore attributable to the sharper increase in attainment for Māori and Pasifika than for European and Asian candidates between 25 and 27. NZ Maori NZ European Pasifika Asian 1 Percentage of participating 11 cohort gaining NCEA Level 1 9 8 7 6 5 4 3 2 1 24 25 26 27 28 29 Figure 26. Percentages of participating 11 Asian, New Zealand European, New Zealand Māori and Pasifika candidates gaining NCEA Level 1 from 24 to 29. All standard errors are less than one percentage point. 29

Figure 27 compares the percentages of participating 12 candidates gaining NCEA Level 2 across the four ethnic groups. New Zealand European and Asian candidates have substantially higher rates of success in gaining NCEA Level 2 during 12 than New Zealand Māori or Pasifika candidates; the differences are quite similar to those evident for NCEA Level 1 presented in Figure 26 above; approximately 13-34 percentage points in favour of the former two ethnicities. Smaller differences in favour of New Zealand European candidates relative to Asian candidates, and to New Zealand Māori candidates relative to Pasifika candidates, are also evident. Although the latter difference appears to contradict Figure 9, which shows a slight difference in favour of Pasifika relative to Māori students in gaining NCEA Level 2 by the end of 12, the apparent reversal is explained by differences in participation rates. Specifically, Figure 9 shows that a higher proportion of the original Pasifika 11 cohort than of the original 11 Māori cohort attained NCEA Level 2 by the end of 12. Figure 26 shows that a higher proportion of the Māori participating cohort than of the Pasifika participating cohort gained NCEA Level 2 during 12. The participating cohort is a more restrictive grouping than the original 11 roll, because it includes only those students with sufficient entries to attain Level 2 in a given year. The participation rate in NCEA Level 2 for 12 Pasifika students is higher than it is for 12 Māori students, such that a higher proportion of all Pasifika students attain Level 2, but the success rate for participating 12 Māori candidates is nonetheless higher than that of participating Pasifika candidates. As is the case for NCEA Level 1 data shown in Figure 26, there is evidence that the large difference between New Zealand European and Asian candidates on one hand, and New Zealand Māori and Pasifika candidates on the other, diminished between 25 and 27. The rate of success for the former two groups has been quite stable over time whereas the success rate for the latter two increased markedly between 25 and 27, and has been relatively stable thereafter. NZ Maori NZ European Pasifika Asian 1 Percentage of participating 12 cohort gaining NCEA Level 2 9 8 7 6 5 4 3 2 1 24 25 26 27 28 29 Figure 27. Percentages of participating 12 Asian, New Zealand European, New Zealand Māori and Pasifika candidates gaining NCEA Level 2 from 24 to 29. All standard errors are less than one percentage point. 3

Figure 28 compares the percentages of participating candidates in 13 gaining NCEA Level 3, across the four ethnic groups of interest. New Zealand European and Asian candidates have substantially higher rates of success in gaining NCEA Level 3 during 13 than New Zealand Māori or Pasifika candidates. There was a small difference in favour of European relative to Asian candidates in 24, but in subsequent years the success rates for these groups were close to identical. The overall success rate of participating candidates increased between 24 and 27, and then fell away somewhat for all ethnicities except Pasifika, which fluctuated downwards in 28 before rising again in 29. As noted previously, the decline for the former three ethnic groups reflects increased retention into 13. The retention of Pasifika students has increased as well, so the underlying success rate for this group might have improved somewhat more than indicated by Figure 28. The difference in favour of Māori candidates relative to Pasifika candidates is consistent with the differences observed for NCEA Levels 1 and 2. As was the case for NCEA Level 2, there is an apparent discrepancy with Figure 11, which shows a slightly higher proportion of the original 11 Pasifika cohort than of the Māori cohort gaining NCEA Level 3 by the end of 13. The explanation for this is the same as that for the apparent discrepancy in the Level 2 data: a higher proportion of all Pasifika than Māori students attain Level 3. The overall participation rate is higher for Pasifika candidates than it is for Māori candidates; however the success rate for participating 13 Māori candidates is higher than that of participating 13 Pasifika candidates. NZ Maori NZ European Pasifika Asian 1 Percentage of participating 13 cohort gaining NCEA Level 3 9 8 7 6 5 4 3 2 1 24 25 26 27 28 29 Figure 28. Percentages of participating 13 Asian, New Zealand European, New Zealand Māori and Pasifika candidates gaining NCEA Level 3 from 24 to 29 All standard errors are less than one percentage point. 31

Figure 29 compares the percentages of 13 students participating in NCEA Level 3, who also gained University Entrance, across the four ethnic groups. Attainment of University Entrance attainment has been quite stable for European candidates, at around 7% of Level 3 participants, although a peak in 27 is evident, with a slight decrease in 28 and 29. The data for European and Asian candidates are close to identical with the exception of 24, in which Asian performance was somewhat lower than European performance. Māori or Pasifika candidates for NCEA Level 3 have fluctuated considerably in attainment of University Entrance, both peaking in 27, and both with somewhat lower rates of attainment in 29 than in 24. NZ Maori NZ European Pasifika Asian 1 Percentage of 13 participants in NCEA Level 3 gaining University Entrance 9 8 7 6 5 4 3 2 1 24 25 26 27 28 29 Figure 29. Percentages of 13 Asian, New Zealand European, New Zealand Māori and Pasifika students participants in NCEA Level 3 gaining University Entrance from 24 to 29. All standard errors are less than one percentage point. 32

Analyses by school decile Figures 3-35 compare the rates of success for participating candidates at schools in low-, mediumand high-decile bands in attainment of NCEA Levels 1-3 and University Entrance in the typical year for each. Data for schools without decile ratings are omitted from these analyses. Figure 3 compares the percentages of participating 11 candidates gaining NCEA Level 1, across low-, medium- and high-decile bands. Performance for all three decile bands increased between 24 and 27 and remained relatively stable thereafter, although there was a slight decrease for candidates at low-decile schools after 27. There are quite consistent differences in favour of high-decile schools relative to medium-decile schools, of between 1 and 15 percentage points, and in favour of medium-decile schools relative to low decile schools of between 13 and 17 percentage points, across the period covered by the data. NCEA Level 1 includes minimum requirements for both literacy and numeracy. Comparisons of the rates at which students at schools in the high-, medium- and low- decile bands attained each of these requirements, independent of whether they attained Level 1 itself, are presented in Figures 31 and 32. Note that these data are based on the 11 roll, rather than on the participating cohort for Level 1 because many students not in the participating cohort for Level 1 nonetheless meet the literacy and numeracy requirements. Decile 1-3 Decile 4-7 Decile 8-1 1 Percentage of participating 11 cohort gaining NCEA Level 1 9 8 7 6 5 4 3 2 1 24 25 26 27 28 29 Figure 3. Percentages of participating 11 students from schools in decile bands 1-3, 4 7 and 8-1 gaining NCEA Level 1 from 24 to 29. All standard errors are less than one percentage point. 33